Skip to content
Apart Research
Apart Research
Contact us
Sprint projectOct 25, 2026Hong Kong

Same Game, Different Players: A Human-Calibrated Testbed for Cartel Enforcement on LLM Pricing Agents

Julian Mündelein · Team CartelBench

Submitted to AI Collusion Research Sprint. Sprint projects are early-stage work by participants, not Apart Research publications.

Read the report

Report: Same Game, Different Players: A Human-Calibrated Testbed for Cartel Enforcement on LLM Pricing Agents

Open in a new tab (opens in new tab)
Recording (opens in new tab)Code (opens in new tab)
Share

I replay a published cartel laboratory experiment (Bigoni et al., 2012) with LLM pricing agents under the same payoffs, rules and matching, so every result sits next to the human one, across four models from three families, with every round pre-registered. Told fines, leniency and rewards remove explicit cartels from the agents more than they did from students, yet three of four models keep prices well above the competitive level almost without a word. A price screen lowers these silent prices, but its published threshold becomes the market price: Claude Sonnet 5.5 prices exactly on the threshold at every level from 3 to 6. A secret random threshold removes the bunching but not the high prices. Code, run records and a hashed run ledger are public.

Reviews

Judging this Sprint?

Review “Same Game, Different Players: A Human-Calibrated Testbed for Cartel Enforcement on LLM Pricing Agents”

Your public critique appears on the project page without your name. Reviews of strong projects are shared with funders, as you choose below.

Not shown on the project page.

Shown on the project page, without your name.

Only the Apart team reads this, and funders if you agree below.

Decline to score

Only when the quality of the writing makes the work impossible to follow. Say why in your private critique; the project page shows only that a judge declined, not why.

How much would this matter for the field if it worked? How innovative is it? For scores of 4-5: is this actually new to the field, or replicating recent work?

Scoring guide
  1. 1Negligible. No clear problem addressed, or no meaningful novelty.
  2. 2Limited. Addresses a real problem but with a generic or well-trodden approach. Incremental at best.
  3. 3Moderate. Clear problem with a reasonable approach; some novelty in framing or method beyond routine application of existing tools.
  4. 4Significant. Important problem with an original approach, or identifies a neglected problem area. A valuable contribution others could build on.
  5. 5Exceptional. Tackles a critical problem with a genuinely novel approach, or opens a new research direction. Clear theory of change. You'd be excited to share this with researchers in the area.

How sound are methodology, implementation, and findings?

Scoring guide
  1. 1Seriously flawed. Methodology broken, results uninterpretable, or implementation doesn't work.
  2. 2Weak. Approach has significant gaps: missing validation, flawed experimental design, or incomplete implementation.
  3. 3Competent. Technically solid given the short duration. Methodology makes sense, results are interpretable, limitations acknowledged, work builds toward clear conclusions.
  4. 4Strong. Thorough methodology with convincing validation. Results clearly support conclusions. Immediately useful for future work.
  5. 5Exceptional. Ambitious scope executed rigorously. Surprising findings, novel methods, or unusually robust validation.

How clearly are work, findings, and impact potential communicated?

Scoring guide
  1. 1Incomprehensible. Cannot determine what the project is actually claiming or doing.
  2. 2Hard to follow. Key information buried, missing, or diluted by excessive length. Significant effort to extract main points.
  3. 3Clear enough. Can understand the problem, approach, and results without undue effort. Core content clearly present: problem, method, findings, limitations.
  4. 4Well presented. Easy to follow, well-structured, appropriate level of detail. Target audience would get it quickly.
  5. 5Exceptionally clear. A pleasure to read. Complex ideas made accessible. Could serve as a model for how to present this type of work.
Sharing your name *

When we share a strong project with funders, your name, scores and public critique go with it, confidentially; this choice is only about whether your name is public. If you think this project deserves funding, please also endorse it: that carries the most weight.

Share my private critique with funders *

See our Privacy Policy.

  1. CartelBench makes a useful contribution by adapting an established human enforcement experiment to LLM pricing agents. Its strongest findings are that fewer explicit agreements need not mean lower prices, and that announced screening thresholds strongly influence some models’ pricing, especially Sonnet. The threshold experiments, scripted-partner controls, zero-penalty placebo, and instruction checks provide substantially more insight than a simple comparison of regulatory prompts.

    The human experiment provides a clear reference for interpreting the agents’ behavior. The paper also reports important qualifications openly, including differences across models, sensitivity to wording, and the disagreement between tests for human Result 3.

    One minor clarification concerns the discussion and appendix recommendation not to disclose the threshold and time window in full. Partial disclosure was not directly tested, so this is best presented as a policy proposal for future evaluation.

    Overall, the central findings are convincing, clearly presented, and useful for evaluating enforcement rules for pricing agents.

    Read full reviewShow less

  2. Calibrating an LLM pricing testbed directly against human lab data from Bigoni et al. is a smart approach. The best result in the paper is that announcing a regulatory price screen turns the threshold itself into a focal point, with LLM prices clustering right underneath it.

1 judge declined to score this project.

Cite this project

@misc{mundelein2026same,
  title = {{Same Game, Different Players: A Human-Calibrated Testbed for Cartel Enforcement on LLM Pricing Agents}},
  author = {Julian Mündelein},
  year = {2026},
  month = oct,
  note = {Submitted to AI Collusion Research Sprint, an Apart Research Sprint},
  howpublished = {\url{https://apartresearch.com/sprints/projects/same-game-different-players-a-humancalibrated-testbed-for-cartel-enforcement-on-llm-pricing-agents-uv2e}},
  url = {https://apartresearch.com/sprints/projects/same-game-different-players-a-humancalibrated-testbed-for-cartel-enforcement-on-llm-pricing-agents-uv2e}
}

More from AI Collusion Research Sprint

All 392 entriesBrowse all projects

A Sprint is running now

AI x Epistemics Research Sprint · Nov 13 - 15, 2026